Upload folder using huggingface_hub

config.json

@@ -1,6 +1,6 @@
 {
   "_attn_implementation_autoset": true,
-  "_name_or_path": "Qwen/Qwen2.5-Coder-7B-Instruct",
+  "_name_or_path": "/tmp/models/tmp9crg3vgi/tmpcrla3rgi",
   "architectures": [
     "Qwen2ForCausalLM"
   ],
@@ -17,6 +17,26 @@
   "num_attention_heads": 28,
   "num_hidden_layers": 28,
   "num_key_value_heads": 4,
+  "quantization_config": {
+    "quant_config": {
+      "offload_meta": false,
+      "scale_quant_params": null,
+      "weight_quant_params": {
+        "axis": 1,
+        "channel_wise": true,
+        "group_size": 64,
+        "nbits": 8,
+        "optimize": true,
+        "round_zero": false,
+        "view_as_float": false
+      },
+      "zero_quant_params": null
+    },
+    "quant_method": "hqq",
+    "skip_modules": [
+      "lm_head"
+    ]
+  },
   "rms_norm_eps": 1e-06,
   "rope_scaling": null,
   "rope_theta": 1000000.0,

qmodel.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6438e1a4a1e8f579ca803cb1c2d1e8b4a421f1f2ded4096ef2d9690d96d52aa5
-size 9114018057
+oid sha256:ab0f3fba59eed7ac22e0c20c035106b78709bdbc92824990c5061d0016376dae
+size 10612246117

smash_config.json

@@ -11,7 +11,7 @@
     "quant_hqq_weight_bits": 8,
     "max_batch_size": 1,
     "device": "cuda",
-    "cache_dir": "/tmp/models/tmp4ijjp52b",
+    "cache_dir": "/tmp/models/tmp9crg3vgi",
     "task": "",
     "save_load_fn": "hqq",
     "save_load_fn_args": {},